I’m working on a project using a Raspberry Pi 4B (4GB RAM) and facing issues when running a Python script that uses DistilBERT and Okt (from the Konlpy library). Below is a simplified version of the code I’m working with:

from transformers import DistilBertTokenizer, DistilBertForSequenceClassification
import torch
from konlpy.tag import Okt

class IntentDetector:
    def __init__(self):
        self.model = DistilBertForSequenceClassification.from_pretrained('distilbert-base-multilingual-cased')
        self.tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-multilingual-cased')
        self.komoran = Okt()

    def recognize_intent_and_destination(self, text):
        input_text = self.preprocess_text(text)
        input_ids = torch.tensor([self.tokenizer.encode(input_text)])
        with torch.no_grad():
            output = self.model(input_ids)[0]
        tokens = self.komoran.pos(input_text)
        destination = self.extract_destination(tokens)
        return tokens, destination

# Test
intent_detector = IntentDetector()
intent_detector.recognize_intent_and_destination("강남역 가는 노선 알려줘")

When I try to run the script, I get a “command not found” error, and the script fails to execute properly. Here’s the key part of the error log:

SyntaxError: invalid syntax
명령어가 잘못됨

I have tried the following steps to troubleshoot:

1. Reinstalled torch, transformers, and konlpy.
2. Increased swap memory on the Raspberry Pi to 2048 MB to handle large model loads.
3. Verified that torch is installed using the correct version for ARM architecture.

Despite these efforts, the issue remains. I suspect it might be related to memory limitations or library compatibility on the Raspberry Pi.

Any suggestions or guidance would be appreciated. Thank you in advance!